IST-3 Takes Another Shot at Extending the 3-Hour Window

1 Comment

It has been well demonstrated that administration of tPA is most effective in the treatment of acute ischemic stroke if given within 3 hours of stroke symptom onset. Furthermore, some experts have even reported this as the standard of care for ischemic stroke since the late 1990s [1].

Using tPA in stroke six hours after onset increases short term morbidity and mortality


It has been well demonstrated that administration of tPA is most effective in the treatment of acute ischemic stroke if given within 3 hours of stroke symptom onset. Furthermore, some experts have even reported this as the standard of care for ischemic stroke since the late 1990s [1]. Within the past five years, the “window” for administration of tPA has been extended to 4.5 hours based solely upon results of the ECASS III trial [2]. Earlier studies attempting to prove efficacy of treating stroke out to 5 or 6 hours proved unsuccessful,  (i.e. Atlantis, 3-5 hours and ECASS I and II, 6 hours), showing increased mortality or no benefit. Despite sufficient evidence that tPA is not beneficial beyond 5 hours, investigators continue to pursue extension of the eligibility window. The most recent attempt to do so is the Third International Stroke Trial (IST-3).

The third international stoke trial (IST-3) was completed to investigate whether a wider range of patients might benefit up to 6 hours from onset of stroke symptoms and on May 23, 2012, the journal Lancet published the results on-line. [3]

The IST-3 is a large, multi-centered, open-label, randomized trial that ran from 2000 to 2011 and enrolled 3035 patents in 156 hospitals across 12 countries, which sought to determine the safety and efficacy of treating stroke with tPA up to 6 hours after onset of symptoms. The results show an increased 7-day morbidity and mortality for patients treated with tPA when compared to placebo controls. At 7 days, an absolute difference of 4% more tPA-treated patients died (Odds Ratio with 95% confidence interval: 1.60 (1.22 to 2.08), p=0.001). This gives a number needed to harm (NNH) of 25, and interestingly enough, this exact NNH was found in the Atlantis trial. In this short-term period, the rate of symptomatic intracranial hemorrhage was 7% among the tPA-treated population and 1% among controls, yielding an absolute difference of 6% (OR with 95% CI: 6.94 (4.07 – 11.8), p< 0.0001) or a NNH of 17. In other words, if you treat 17 patients with stroke up to 6 hours from onset, one more will have a symptomatic intracranial hemorrhage. If you treat 25 patients, one more will die.



Interestingly, this difference in mortality went away when the population was followed out to 6 months. The 6 month mortality was identical for both groups at 27%. (OR with 95%CI 0.96 (0.80 – 1.15), p = 0.672. This developed because 55 more patients died in the control group between 7 days and 6 months, an absolute difference of 4%. The deaths in the control group neutralized the impact of the early deaths seen from tPA. However, it is unclear what the cause of death was for the patients in the control group. Perhaps, they were not related to a lack of tPA administration or even cerebrovascular disease.

The primary outcome for the trial was being alive and able to live independently at 6 months based upon the Oxford Handicap Score (OHS), which is a commonly used variant of the modified Rankin score [4]. Patients with an OHS score of 0, 1 or 2 were classified as being independent where OHS of 0 means no change in lifestyle, OHS 1 means no interference in lifestyle, and OHS 2 means some lifestyle restrictions but able to look after one’s self [5]. For this primary outcome, there was no difference between the groups (37% for treated subjects vs. 35% for controls, OR with 95% 1.13 (0.95 – 1.35), p = 0.181. 

Unfortunately, a press release by the National Stroke Association published in HealthDay News and republished in a number of news outlets seems to have mischaracterized the results of this study. That report says, “After six months, slightly more of those who got tPA – 37 percent vs. 35 percent – were alive and independent.” [6] In fact, the rate of being alive and independent was the same between both groups and demonstrated no statistically significant difference in independence at 6 months. The statement that was published is misleading as it implies that there is a long-term benefit at 6 months when tPA is administered up to 6 hours after symptom onset, although this is not actually the case. What was not mentioned is that there was a statistically significantly difference in 7 day morbidity and mortality.  As reported in HealthDay News, “there was a slight benefit for those who received tPA within 6 hours.” In fact there was no added benefit when tPA was given within 6 hours and was associated with slightly worse outcomes at 7 days post-tPA.


Overall, this is a well-designed study providing the strongest possible evidence. It is randomized, multi-centered, and placebo-controlled. The treatments the patients received were open-label and the 7 day measurement was taken with knowledge of the treatment group, but the study designers masked these treatments at the 6 month follow-up.

Very few patients were lost to follow-up (11 in the treatment group and 13 in the control group). One study limitation was that there was a problem with the study’s power analysis. Originally powered to detect a 3% difference in outcome at 6 months, the study sought to enroll 6000 patients. In 2007, after 7 years of data collection, it was clear that study would not be able to enroll this many patients and another power calculation was made. They determined that 3100 patients would have the power to detect a 4.7% difference in outcomes.

 The study started in 2001 and the recalculation of the power occurred in 2007. During this period of time, several studies demonstrating efficacy of tPA of stroke within 3 hours were published. From this and other studies, we know that patients tend to present late when they develop stroke symptoms. In fact, 45% of patients in IST-3 were enrolled after 4.5 hours.

This could introduce length of time bias in favor of patients presenting further out from symptom onset. We speculate that patients presenting after 4.5 hours were more likely to become subjects of this study, at least those enrolled after the power recalculation in 2007. Since the “acceptable window” for tPA in stroke became 4.5 hours as of ECASS III in 2008, patients presenting more than 4.5 hours out would have two options: no therapy or enroll in IST-3. 

Despite our support for the methodological design of IST-3, the interpretation of the results the investigators, however, is puzzling and seems to represent spin, as opposed to a true reflection of the study results. While the IST-3 aims to assess whether a wider range of patients might benefit from tPA up to 6 hours from stroke onset, it does not provide strong evidence in favor of this hypothesis and in fact may slightly increase 7 day morbidity and morbidity. This study demonstrates what has been shown previously; tPA is most beneficial if given within 3 hours and that it is likely harmful beyond 4.5 hours. Further study of the expansion of the tPA eligibility time frame seems unwarranted and may subject patients to unnecessary risk. In addition, such studies have raised questions about the safety and efficacy of tPA. Thus, larger scale studies, such as IST-3 could be performed to also include patients that present within 3 hours, which could definitely answer any remaining questions about the safety and efficacy of tPA.

Before you decide to order tPA, you may want to consider IST-3, which could not demonstrate a statistically significant improvement in 6-month functional independence, but did result in a NNH of 25 at 7 days. (NINDS rt-PA
Stroke Study Group. Tissue plasminogen activator for acute ischemic stroke. The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. N Engl J Med. 1995; 333: 1581–1587.)


  • Hacke W, Kaste M, Bluhmki E, Brozman M, Dávalos A, Guidetti D, Larrrue V, Lees KR, Medeghri Z, Machnig T, Schneider D, von Kummer R, Wahlgren N, Toni D; ECASS Investigators. Thrombolysis with alteplase 3 to 4.5 hours after acute ischemic stroke. N Engl J Med. 2008; 359: 1317–1329
  • IST-3 collaborative group. The benefits and harms of intravenous thrombolysis with recombinant tissue plasminogen activator within 6 h of acute ischaemic stroke (the third international stroke trial [IST-3]): a randomised controlled trial. Lancet. 2012
  • Bamford J, Sandercock P, Dennis M, Burn J, Warlow C. A prospective study of acute cerebrovascular disease in the community: The Oxfordshire Community Stroke Project-1981-86. 2. Incidence, case fatality rates and overall outcome at one year of cerebral infarction, primary intracerebral and subarachnoid hemorrhage. J Neurol Neurosurg Psychiatry 1990: 53: 16-22.
  • Bamford J, Sandercock P, et al. Letter to the Editor: Interobserver agreement for the assessment of handicap in stroke patients. Stroke. 1989; 20: 828 (Table 2).


1 Comment

  1. Charles A. Pilcher MD FACEP on

    “Spin”??? What a concept! Wonder who might have skin in this game, and what good “spin” costs these days.

Leave A Reply